[Web-bot-auth] Re: Interest in the human-principal layer above bot authentication

[Songbo Bu <[e-mail]>]

Nenad, all,

I reproduced the suite-transition pair independently against the exact
base commit:

46ad663b90805a2e526ef3cd28c3f70762883125

The list-delivered patch had CRLF line endings and an mbox-escaped
first line beginning with >From. Normalizing only those transport
changes restored the declared SHA3-256:

ac2706b4abf64515755df0d00865f57a55504345bdb66f64ecad9a900d0b0830

The normalized patch applied cleanly with git am in a separate clone.
The reproduced results were:

- Python verifier: CONFORMANT, 46 vectors, both verdicts, 10 reject
reasons, 9 kinds;
- TypeScript cross-check: all 46 decisions agreed;
- differential harness: 258 cases, including 212 off-corpus mutations,
with zero divergence; and
- regeneration: all 46 vectors and the manifest were byte-identical,
with a clean worktree afterwards.

My technical reading is that the pair is discriminating. The positive
vector binds the retained prefix under the digest suite in force when
that prefix was written. The negative twin supplies the
successor-suite digest of the same canonical prefix, so a verifier
that retroactively re-digests history computes the claimed value and
is caught.

The evidence boundary remains narrow: this reproduces the pair against
the two pinned verifier paths in the shared tree. It does not by
itself establish adoption, independent-party implementation, or the
broader truth or completeness of the recorded events.

Subject to stabilizing the numbering after the predecessor vectors
land, I support adding this pair to the shared corpus.

Best,
Songbo